Despite publication yesterday of a report which has been generating headlines that immigration causes unemployment, IPPR’s Matt Cavanagh argues that this is far from a ‘gotcha’ moment in the on-going immigration debate.
The impact of immigration on unemployment has been central to the political debate in recent years, especially since unemployment started rising after the financial crisis. Many people – including Coalition ministers – are strongly wedded to an intuitive picture in which the two are locked together in a zero-sum game: migrants compete for jobs with residents; therefore more migrants must mean fewer jobs for those who are already here. But economists have been more sceptical.
Yesterday saw the publication of two important new reports on this issue. (We can disregard Monday’s report from the pressure group MigrationWatch, the patent inadequacy of which was rehearsed here.)
The first report is from the Migration Advisory Committee (MAC), a committee of economists set up to advise the government on immigration policy. The MAC confirms the finding from previous studies that immigration does have an effect on wages: it finds no overall effect, but a negative effect on lower-paid workers, offset by a positive effect on higher-paid workers.
What will make the headlines, however, is that for the first time ever in a government-sanctioned report, the MAC finds an ‘association’ between immigration and employment. In particular, the MAC estimates that: ‘an extra 100 non-EU migrants are initially associated with 23 fewer native people employed’. Many will seize on this as a ‘gotcha’ moment, vindicating their intuitive view about the link between immigration and unemployment. However, a number of cautionary points need to be made.
First, this relates only to non-EU migrants. It isn’t in any way a vindication of MigrationWatch and others who were arguing in the media yesterday about migrants from eastern Europe. Indeed, the MAC explicitly states that it finds no association between EU migrants and unemployment.
Second, it relates only to short-term migrants. For those who stay longer than five years, the MAC again finds no association between migration and employment. This is significant in relation to the government’s proposals to severely restrict the opportunities for working migrants to stay longer than five years: it means the government cannot cite economic or job-related arguments in favour of this policy. (IPPR has criticised the government’s proposals in this area here.)
The combination of these two points means that, while the figure of ‘100 migrants displace 23 British workers’ will doubtless get the attention, the MAC’s overall finding, covering EU and non-EU migrants, and those who stay over five years, is that the 2.1 million additional foreign workers in 2010 compared with 1995 are ‘associated with’ 160,000 fewer resident workers: a ratio of 1 in 13 rather than 1 in 4.
Third, hidden in the detail of the MAC report is an important distinction between periods of economic growth and economic downturns. The MAC has chosen to present its results for the whole period (from 1995 to 2010). But if it had chosen instead to separate the two – for example, to look at the period 1995–2008, and then separately at 2008–2010 – the same results would have yielded an ‘association’ between immigration and unemployment in a downturn, but not in periods of economic growth(see table 4.1 on p63).
Fourth and last, the MAC is clear that it is not claiming to have found a causal link – it is merely ‘an association’. It also admits to some important limitations on the robustness of its results (readers of a statistical bent might like to look at paragraph A67 on pp116–117).
The other report published today, by NIESR, looks at both EU and non-EU migrants. Crucially, moreover, it is the first major study to use data from national insurance numbers, rather than the survey data (including the Labour Force Survey) that is used in most immigration analyses, including the MAC’s report. This makes the NIESR study more robust, and also allows us to drill down to different sectors and parts of the country. NIESR’s report finds no link between immigration and unemployment – contrary to the MAC’s report, but in line with most previous studies.
What should we conclude? The MAC report is an important study – but has important limitations; and despite its government imprimatur, it is only one study among many. Opponents of immigration should reflect on the finding that any ‘association’ between immigration and lower employment for residents does not last if migrants stay, and on the fact that the MAC find no ‘association’ in relation to EU migrants, despite the fact that the intuitive argument seems to be strongest in relation to eastern European migrants – a point which MigrationWatch emphasised heavily yesterday.
Britain clearly has a major problem with unemployment, including a problem with youth unemployment which predated the financial crisis. But its causes are too complex to be reduced to blaming immigration, just as the effects of immigration on the labour market are too complex to be reduced to endlessly repeated headlines about ‘foreigners taking all the jobs’. This important debate should, and will, continue. This week’s reports are far from conclusive in proving that cutting immigration will make a significant contribution to reducing unemployment, let alone be the silver bullet which many believe it is – though we can expect that to be how it will be reported in many of this week’s papers.
Please read our comments policy before posting.
This article was originally published at the IPPR on 10 January.
Matt Cavanagh – Associate Director for UK Migration Policy, IPPR
Matt has extensive broadcast media experience including on BBC TV and radio, Sky and others. He writes for Prospect and Spectator magazines, and has published online articles for the New Statesman, the Independent, the Guardian’s Comment Is Free, and elsewhere. Matt previously worked as a special adviser for the Labour government between 2003 and 2010, in the Home Office, the Treasury, the Ministry of Defence, and finally in the Number 10 Policy Unit from 2007 to 2010.